The VISIBLE LEARNING Approach to Unleashing Education Improvement - Cognition Education Group
←
→
Page content transcription
If your browser does not render page correctly, please read the page content below
The VISIBLE LEARNING®
Approach to Unleashing
Education Improvement
A JOINT PUBLICATION
Getting to
G.O.L.D.
The VISIBLE LEARNING®
Approach to Unleashing
Education Improvement
Arran Hamilton | John Hattie
A JOINT PUBLICATION
FOR INFORMATION: Copyright © 2021 by Corwin Press, Inc.
Copyright © 2021 by Corwin Press, Inc. Visible Learning® and Visible Learning+
Corwin
are registered trademarks of Corwin Press, Inc. All rights reserved. Except as
A SAGE Company permitted by U.S. copyright law, no part of this work may be reproduced or distributed
2455 Teller Road in any form or by any means, or stored in a database or retrieval system, without
Thousand Oaks, California 91320 permission in writing from an authorized representative of Corwin Press, Inc.
(800) 233-9936
When forms and sample documents appearing in this work are intended for
www.corwin.com reproduction, they will be marked as such. Reproduction of their use is authorized for
educational use by educators, local school sites, and/or noncommercial or nonprofit
SAGE Publications Ltd. entities that have purchased the book.
1 Oliver’s Yard
All third-party trademarks referenced or depicted herein are included solely for the
55 City Road
purpose of illustration and are the property of their respective owners. Reference to
London EC1Y 1SP these trademarks in no way indicates any relationship with, or endorsement by, the
United Kingdom trademark owner.
ISBN 978-1-0718-1998-2 (web PDF)
SAGE Publications India Pvt. Ltd.
B 1/I 1 Mohan Cooperative Industrial Area
Mathura Road, New Delhi 110 044
India
SAGE Publications Asia-Pacific Pte. Ltd.
18 Cross Street #10-10/11/12
China Square Central
Singapore 048423
Acquisitions Editor: Ariel Curry
Editorial Assistant: Eliza Erickson
Production Editor: Laureen Gleason
Copy Editor: Integra
Typesetter: C&M Digitals (P) Ltd.
Proofreader: Sarah J. Duffy
Graphic Designer: Lysa Becker
Marketing Manager: Kerry Garagliano
DISCLAIMER: This paper may direct you to access third-party content via web links, QR codes, or other scannable
technologies, which are provided for your reference by the author(s). Corwin makes no guarantee that such third-party
content will be available for your use and encourages you to review the terms and conditions of such third-party content.
Corwin takes no responsibility and assumes no liability for your use of any third-party content, nor does Corwin approve,
sponsor, endorse, verify, or certify such third-party content.
Contents
About VISIBLE LEARNINGiv
About Corwin iv
About the Authors v
Acknowledgmentsv
Introduction1
Part 1 Overview of the Implementation Challenge 3
1.1 Why Is Education Reform Hard? 3
1.2 Toward an Implementation Science for Education 7
Part 2 The VISIBLE LEARNING G.O.L.D. Model 10
Overview 10
Phase 1 Goal Hunt 12
1.1–1.3 Hunting for an Education Challenge Worth Solving 13
1.4 Map the Causal Drivers 15
1.5 Set the Quantifiable Improvement Goal 16
Phase 2 Opportunity Sift 19
2.1 Reverse Engineer from Causal Drivers to Opportunity Sketches 20
2.2 Develop a Theory of Improvement 22
2.3 Iterating Your Theory of Improvement 22
2.4 Develop Your Success Map 29
Phase 3 Lift Off 32
Phase 4a Double Back 34
Phase 4b From Double Back to Double Up 38
Conclusion41
References42
About VISIBLE LEARNING
In 2008, Professor John Hattie published Visible Learning, a synthesis of more than 800 meta-studies
covering more than 80 million students. The book revealed what education variables have the biggest
impact on learning and created a new mindset that has swept educators around the world. Applying the
Visible Learning® methodology means that students are taught to know what they need to learn, how to
learn it, and how to evaluate their own progress. Using the Visible Learning approach, teachers become
evaluators of their own impact on student learning. The combination causes students to drive their own
learning. Since 2008, Professor Hattie has teamed with highly influential educators to expand the Visible
Learning canon with books, including Visible Learning into Action, Visible Learning for Teachers, Visible
Learning for Mathematics, and Visible Learning for Literacy.
Visible Learning+ is the model of professional learning that takes the theory of Hattie’s research and puts
it into a practical inquiry model for teachers and school leaders to ask questions of themselves about the
impact they are having on student achievement. Visible Learning+ is a result of the collaboration between
Professor John Hattie and Corwin to help educators translate the Visible Learning research. Through a
global network of partners, Visible Learning+ professional learning is implemented in over 20 countries in
North America, Europe, and the Pacific.
Learn more at www.visiblelearningplus.com
About Corwin
Corwin, a SAGE Publishing company, was established in 1990, first as a professional book publisher, now
as a full-service professional learning company, offering professional development solutions on the topics
that matter most and the delivery methods that work best to achieve a school or district’s objectives.
Its many resources range from a library of over 4,000 books to on-site consulting to online courses and
events. At the heart of every professional learning experience is the book content and author expertise
that have made Corwin the most trusted name in professional development.
Learn more at www.corwin.com
iv
About the Authors
r. Arran Hamilton is Group Director of Education at Cognition Education. His early
D
career included teaching and research at Warwick University and a period in adult and
community education. Dr. Hamilton transitioned into educational consultancy more
than 15 years ago and has held senior positions at Cambridge Assessment, Nord Anglia
Education, Education Development Trust (formerly CfBT), and the British Council. Much
of this work was international and has focused on supporting Ministries of Education
and corporate funders to improve learner outcomes. He currently oversees the design
and delivery of Cognition Education’s portfolio of education improvement programs,
with live projects in the UK, Australia, New Zealand, and the Pacific Islands.
Professor John Hattie is Laureate Professor at the Melbourne Graduate School of
Education at the University of Melbourne and Chair of the Australian Institute for
Teaching and School Leadership. His areas of interest are measurement models and
their applications to education’s problems, and models of teaching and learning. He
has published and presented over 1,000 papers, supervised 200 thesis students, and
published 31 books, including 18 on understanding and applying the Visible Learning®
research.
Acknowledgments
We would like to thank Dr. Brian Hincho, Dr. Shaun Hawthorne, Mary Sinclair, and Christophe Mullings
from Cognition Education; Ariel Curry and Julie Smith from Corwin; and Distinguished Professor Viviane
Robinson and the eight anonymous reviewers who commented on an early draft of this paper. Any remain-
ing errors or omissions are entirely our own.
v
Introduction
Achieving an Olympic Gold is the pinnacle of distilling these into 270+ separate influences on
success for world-class athletes. But running, student achievement. These data are now publicly
say, 100 meters in under 11 seconds is fiendishly available for all on the Visible Learning MetaX, avail-
difficult. It requires elite performance and get- able at http://www.visiblelearningmetax.com/.
ting to gold requires both the what and the how.
Like the Olympic coaches and their athletes, this
The what consists of the burgeoning research on
means educators now have access to a good
the “running economy”: the efficiency of different
(enough) compass to point (broadly) in the direc-
running gaits, training regimes, diets, and equipment/
tion of education gold. We know what works best
attire to enhance performance on the track.
or rather what has worked well previously in a range
The how consists of athletes’ approaches to imple- of different contexts. Educators can use these data
menting these learnings, in order to give them crit- to make forward-looking probability estimates, or
ical edge. It involves the support of a coach who bets, about what is more (or less) likely to work well
can advise on technique, and often the method in their contexts in the future.
of implementation is highly scientific. It starts with
However, one of the key remaining challenges for
the analysis of athlete-specific data and leads to
educators is that tricky question of how. An unan-
the development and testing of hypotheses (or
ticipated outcome of the Visible Learning project is
hunches) about how to improve, based on review
that (in some contexts) it has encouraged educators
of global research. This involves large doses of trial
to simply pick the interventions with the highest
and error because what works best for Usain Bolt
effect size—irrespective of whether these address
does not necessarily work best for Tyson Gay or
education challenges they genuinely have in their
Florence Griffith-Joyner.
local context. This is the equivalent of a sports
What + How = Impact. coach blindly changing their runner’s footwear—
based on the global data, rather than by looking
In our world of education, there are many parallels. closely at the specific areas of improvement that
Like Olympic coaches, every educator and educa- their athlete needs.
tion system seeks gold for their learners. Although
in this case, gold is at least a year’s growth for a A second conundrum that educators then face is
year’s teaching input. how to implement the identified approaches with
fidelity. Too often, this has proved fiendishly diffi-
Educators are also becoming increasingly adept at cult. Either the original program designers provided
drawing on the burgeoning research on what works no explicit guidelines on how or those guidelines
best to enhance their performance. Through the just didn’t seem to work quite as intended in the
Visible Learning research, we have harvested and syn- new context. In fact, one of the most pressing prob-
thesized the findings of more than 96,000 research lems we face in education is the challenge of repli-
studies involving more than 300 million students: cation, particularly at scale.
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
1
We do not yet have a well-developed science • Goal Hunt—agreeing on what needs to be
of improvement that enables educators to improved
systematically
• Opportunity Sift—agreeing on the mechanisms
that provide opportunity for improvement;
• identify education challenges worth solving;
• develop (or select) interventions that are likely to • Lift Off—implementing the improvement; and
have impact;
• Double Back—measuring the impact and agreeing
• implement with fidelity; on where to go next, so that, ideally, we can also
• evaluate to determine whether there has been Double Up to scale the impact.
impact; and
• iterate to enhance impact.
This has most similarity to the idea that “teachers
This paper is about that challenge of selection and are to DIIE for”; that is, educator impact comes from
implementation. Our key purpose is to sketch out excellent Diagnoses, choosing high-probability
some protocols to enhance the probability that the Interventions, ensuring Implementation fidelity and
quest for gold will actually lead to gold. appropriate dosage, and conducting Evaluations
of the impact on the learning lives of students (see
In keeping with the theme of this collection of Hattie & Zierer, 2018).
papers, we call our implementation approach the
G.O.L.D. Method. As we will unpack, this involves On our journey, we will traverse the following
four phases: waypoints:
Part 1 Overview of the Implementation Challenge
Why Is Education Reform Hard? This section celebrates the wonderful success that governments have made in
scaling education for all but laments that progress is slowing and that improvement is getting ever harder.
Toward an Implementation Science for Education, which surveys a range of promising approaches that are
currently being piloted in education, including Deliverology and Spiral of Inquiry. We argue that there are golden
nuggets in each that could benefit from being extracted and re-blended.
Part 2 The Visible Learning® G.O.L.D. Model
Overview of G.O.L.D. This introduces the key phases and stages.
Goal Hunt. This phase outlines processes for the identification of education challenges worth fixing. In other
words, this is about hunting for the right issues to address.
Opportunity Sift. This suggests protocols for sketching, building, and validating theories of improvement. In
other words, this is about designing effective interventions.
Lift Off, which is where interventions are implemented.
Double Back. This is about evaluating the effectiveness of what you have done to identify how you can double
up the impact.
Finally, we include our observations about moving from the Double Back phase to Double Up, that is, to achieve
impact at scale.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
2
PART 1
Overview of the Implementation Challenge
1.1 Why Is Education Reform Hard?
Let’s start with some good news. The last 170 years • The Access Challenge. There are still 263
has, educationally speaking, been a resounding million children (one in five adolescents and
success. When school systems first emerged, the disproportionately girls) around the globe that are
trailblazers were like bush runners. They had no not currently in school (UIS, 2019).1 This is more
path, no world-class coaches, and no custom-made than the combined child-age population of the
“running shoes.” Despite this, those early pioneers United States, Canada, the United Kingdom, the
ran like the wind and quickly massified education. Netherlands, Scandinavia, Australia, and
New Zealand.
Since the 1850s, governments around the world
have moved from an era of no publicly funded uni- • The Equity Challenge. In developed countries,
versal education to single-room primary schools, where access to schooling is universal, education
and then to the complex modern schools we see in outcomes are still too much of a lottery—with an
many parts of the world today. Governments have average of 14% of students in OECD countries
literally pumped hundreds of trillions of dollars into failing to graduate high school (OECD, 2019,
this endeavor. Schools have been built. Teachers Secondary graduation rate indicator). This
have been identified, trained, and hired. And nongraduation rate includes 21% of boys in the
according to World Bank data, currently 89% of United Kingdom, 18% of boys in the United States,
children across the globe have had access to some and 17% of boys in Australia.
form of schooling (UNESCO UIS, 2018).
• The Future Skills Challenge. Employers and
In many countries, education is the single biggest educators regularly lament that the education
area of public investment, with governments now col- system seems to have been built for a bygone
lectively spending more than USD $3.46 trillion per factory era and that it does not equip young people
annum on both basic and higher education (Hattie & with the skills they need for the world of today, let
Hamilton, 2020). The global spread of schooling has alone tomorrow.
been one of the greatest successes of our era.
The global spread of schooling has been one of In many countries, policymakers appear very con-
the greatest successes of our era. scious of the fact that their collective investments
in education are not bearing enough fruit. There is
Despite these successes, many stakeholders remain also growing concern about the potential for mass
unhappy because of the following hurdles: unemployment as algorithms become primed
1
Many of these children have previously had some access to primary-level education.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
3
to undertake the more “routine” tasks like stock There can be many positive consequences of some
picking, pathology, and legal review (Ford, 2017). of this, but GERM is premised on the belief that it is
The historian-cum-futurologist Yuval Noah Harari optimal to primarily focus on the levers external to
(2019) talks with great seriousness about the the school and classroom. The opposite has been
likely emergence of a useless class whose current equally as fraught—the belief in a school leader or
employment activities could be entirely automated teacher’s right to autonomy to choose their own
but who lack the creative and critical thinking skills teaching adventure, the focus on how to teach
to reinvent themselves. Indeed, the one thing our rather than on the impact of this teaching, the
machine colleagues can’t (yet) do is emulate the denial of expertise among educators, and the false
uniquely human skills of curiosity, creativity, col- belief that all educators are equal in their impact so
laboration, critical reflection, and communication leave them alone.
that will likely be the key to success for future
However, as nations increasingly use international
generations (Fullan & Scott, 2014; Jefferson &
rankings like the Programme for International
Anderson, 2017).
Student Assessment to measure the comparative
virility of their respective national education sys-
Recognizing the risks, education policymak-
tems, we are witnessing greater convergence. The
ers across both the developing and developed
standardization of the measuring instrument seems
worlds keep themselves busy with the task of
also to be encouraging the standardization of pol-
designing and implementing education reform.
icy prescription.
That playbook of reform varies from time to time
and place to place, but as Pasi Sahlberg laments, As one of us has argued elsewhere (Hattie, 2015),
it is increasingly looking like a Global Education many of these system-level reforms are largely
Reform Movement (GERM) (Sahlberg, 2015), barking up the wrong tree or, rather, they don’t
which is a standardized toolkit of policy inter- bark high enough. National education systems are
ventions that centers on the initiatives tabulated a composite of five interlocked layers, as shown
in Table 1. in Table 2.
Table 1 The GERM Model of Education Improvement
Policy Prescription
Standardized Teaching • National teaching standards
and Learning
• National curriculum
• Common data standards
Market-Based Reforms • School inspectorate
• School league tables
• Teacher performance-related pay
• Parental choice
• School autonomies
Test-Based Accountability • National standardized testing, linked to market-based reforms
Focus on Literacy and • Increasing teaching hours for math, reading, writing, and science
Numeracy
Source: Sahlberg, 2015.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
4Table 2 The Five Layers (or Levels) of Education Improvement
Layer Description
Level 1: Big Picture • Purpose of education, e.g., basic human right, economic productivity, and equal
opportunities
• National education challenges, e.g., ensuring all children get a quality education,
increasing literacy rates, reducing truancy, zero tolerance to bullying, and unlocking
creativity and unique human skills
Level 2: Education The policy instruments that are used to effect change at Levels 3 and 4:
System Surface
• Education funding, facilities, curriculum, school governance structures, assessment
Structure
system, teacher recruitment and training processes, data monitoring systems,
class size, length of school day, number of minutes devoted to specific subjects,
budget for teacher professional development, and level of centralization versus
decentralization
Level 3: Education What teachers actually do in classrooms and what they believe (espoused theories vs.
System Deep theories in practice):
Structure
• Teachers’ theory of learning and their role in it, how they interact with
learners, how they provide empathy, how they give and receive feedback,
how they gather evidence and reflect on their strengths and areas for development,
how they collaborate with each other, and the micro-pedagogies they employ
Level 4: The Learner • What and how children learn; the facts, values, and skills that are transmitted and the
fidelity of transmission (or the neurons that fire and wire together)
Level 5: The Legacy • The level of long-term contribution that schooling makes in equipping young people
Effect with skills for life versus the impact of other formal and informal mechanisms for
learning, e.g., environmental context, family, peers, voluntary associations, and TV
• The fadeout/degradation of learning over time, i.e., how much of what children
learn in school is used and useful for life?
The implicit assumption behind many current and ecosystem and life outcomes. And the assumed
recent national education reform initiatives is that direction of that relationship is school transmitting
there is a direct causal chain or golden thread that to life rather than vice versa.2
is cast down from Level 1 and that sways vigorously
at Level 5. In other words, the assumption is that Our sense is that what happens at Level 3 has
policymakers set goals; these goals are converted a very strong impact on Level 4 and that this, in
into uniform instruments; these instruments, once turn with various out-of-school influences, has
implemented, significantly impact, and for the bet- some impact in laying appropriate foundations for
ter, what teachers do in classrooms. It is further Level 5. However, we are not at all convinced that
assumed that this change in teacher behaviors then much of what happens at Levels 1 and 2 has any
directly results in children learning faster, harder, significant impact on Levels 3–5. Our contention is
and better and that there is a causal relationship that both segments operate, too often, in parallel
between what goes on in this whole schooling universes.
2
Not all educational theorists hold with this view, however, from Dewey to Illich to Beane to Yong Zhao, to name a few.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
5There are many explanations for why the growing differently, multiplied by the number of change
global body of research about what works trans- agents that can help them on their journey. No
lates into limited impact in schools and classrooms. wonder that Larry Cuban (1998) in his analysis of
First, education is often a highly politicized area the impact of John Dewey’s progressive educa-
of public service delivery (Nordstrum et al., 2017). tion movement in the United States concluded
This means that new programs can sometimes be that even at its peak no more than 25% of teach-
implemented based on political or ideological con- ers had fully converted to using student-centered
siderations and without any clear linkage to the evi- approaches. The majority did not change anything
dence on what works or any evaluation of actual at all and many of those teachers who said they
impact. In the United States, for example, federal were progressivists employed hybrid approaches
funding for education programs has rarely been that retained many of the features of their preexist-
tied to impact data (Slavin, 2002). ing teaching practice.
Second, even where appropriate and research-
based interventions have been identified, achiev- The success of any educational reform is directly
ing implementation with fidelity is extremely proportional to the number of teachers who are
difficult (Berman & McLaughlin, 1975; Cuban, willing to question their own approaches and to
1993; Darling-Hammond & Snyder, 1992; Elmore look seriously at ways of doing things differently,
& McLaughlin, 1998; Rivlin & Timpane, 1975; multiplied by the number of change agents that
Robinson, 2018). Interventions and approaches can help them on their journey.
that researchers agree on paper should be effec-
tive often do not produce the intended results or
To overcome these challenges, we need to widen
significantly change what educators do in the class-
our focus from what works to how we can make it
room (Cohen & Moffitt, 2009; Correnti & Rowan,
work. We need to ask what works best for whom
2007; Rowan, Correnti, Miller & Camburn, 2009).
and under what conditions; or why it worked earlier
Ultimately, the success of any educational reform and how we can make it work again. In short, we
is directly proportional to the number of teachers need a science of improvement (including effec-
who are willing to question their own approaches tive implementation of those improvements) for
and to look seriously at ways of doing things education.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
61.2 Toward an Implementation
Science for Education
In the parallel field of health, implementation new approach aligned with health care
science has had a reasonably long history. Methods professionals’ worldviews, it was more likely
for effective rollout and scale-up of health reforms to be implemented with enthusiasm (Aarons,
have been iterated and improved since the 1960s Green, & Miller, 2012). Ergo, implementation
(Pressman & Wildawsky, 1984; Wandersman et al., must tackle beliefs.
2008). These approaches emerged because empir-
3. Practitioners should not have sole
ically tested health intervention programs were not
responsibility for implementing research-
having the desired impact, once implemented in
based interventions. Instead, accountability
primary care settings.
for success also lies with the researchers and
program developers, who should work in
The research in health care suggested tandem (Meyers et al., 2012).
In our business of education, implementation sci-
1. The actual process of implementation ence has had a relatively short history. The Hand-
strongly influenced whether there book of Implementation Science for Psychology in
was impact (Kelly, 2012; Meyers Education, edited by Kelly and Perkins, was pub-
et al., 2012). lished in 2012. This was the first weighty tome on
the psychology of education implementation.
2. There is a powerful relationship
between the beliefs and values of health There are also several promising improvement
care practitioners and the outcome methodologies for education that are being piloted
of implementation; that is, when the in a range of education contexts (Table 3).
Table 3 Overview of Some Promising Education Improvement Methodologies
Methodology Description Reference
Deliverology Developed by Sir Michael Barber, this approach Barber, M., Kihn, P., & Moffit, A.
focuses on scaled implementation of an agreed (2011). Deliverology 101: A field
approach, with fidelity. The framework is intended guide for educational leaders.
to be generic (i.e., can be used within and beyond Thousand Oaks, CA: Corwin.
education) but has been employed in educational
contexts in the United States.
Verdict: Strong on governance, implementation, and
scalability processes but limited focus on protocols for
selecting appropriate interventions or for engaging
with stakeholders’ beliefs and values. More suitable
for top-down change.
(Continued)
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
7(Continued)
Methodology Description Reference
Positive Centers on grassroots exploration and fact finding to LeMahieu, P. G., Nordstrum, L., &
Deviance identify positive outliers or “deviants” to whatever Gale, G. (2017). Positive deviance:
issue stakeholders seek to resolve. The idea is to Learning from positive anomalies.
catalogue positive deviant behaviors that can be Quality Assurance in Education,
replicated and scaled up. 25(1). https://doi.org/10.1108/
QAE-12-2016-0083
Verdict: Strong on innovating problem solving
by scaling up pockets of effective practice, but Pascale, R., Sternin, J., & Sternin,
limited focus on how stakeholders select their issue, M. (2010). The power of positive
how they should scale up, or how to address the deviance: How unlikely innovators
fact that many instances of positive deviance are solve the world’s toughest
nonreplicable. problems. Boston, MA: Harvard
Business Press.
Spiral of Developed by Helen Timperley, Linda Kaser, and Timperley, H., Kaser L., & Halbert, J.
Inquiry Judy Halbert, the Spiral of Inquiry provides frontline (2014). A framework for transforming
educators with six subroutines to enhance student learning in schools: Innovation
learning outcomes: scanning, focusing, developing a and the spiral of inquiry. Victoria,
hunch, learning, taking action, and checking. Australia: Centre for Strategic
Education.
Verdict: Strong on identification of areas for
improvement and the development, implementation
of local action research projects to implement and
measure improvement and on privileging stakeholder
beliefs. Less focus on specific implementation
processes or on scaling up.
Agile for Range of approaches, including Simon Breakspear’s Breakspear, S. (2017). Embracing
Education Teaching Sprints, which have been adapted from agile leadership for learning: How
the generic Agile model and applied to educational leaders can create impact despite
settings. growing complexity. Australian
Educational Leader, 39(3), 68–71.
Verdict: These frameworks have similar focus and
benefits to the Spiral of Inquiry model and are
valuable process drivers for school or departmental
professional learning communities to undertake fast
improvement cycles. They may have less value in
driving system-wide reform.
Reduce This is a body of school improvement research Robinson, V. M. J. (2018). Reduce
Change to and processes developed by Viviane Robinson. It change to increase improvement.
Increase proposes a highly selective improvement focus and Thousand Oaks, CA: Corwin.
Improvement offers protocols for coaches to engage with and
understand stakeholder beliefs around change.
Verdict: Wonderfully strong on “less being more”
and on mechanism for engaging with rather
than bypassing educators’ theories of action and
their beliefs. Less coverage on the processes
of implementing, evaluating, and scaling up
improvement.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
8Methodology Description Reference
Hexagon Tool A six-part planning tool for schools to identify local Blase, K., Kiser, L., & Van Dyke, M.
needs and then evaluate prebuilt programs and their (2013). The hexagon tool: Exploring
suitability for the local context. Focus areas are need, context. Chapel Hill, NC: National
fit, resources, evidence, readiness, and capacity. Implementation Research Network,
FPG Child Development Institute,
Verdict: A useful thinking tool to support individual
University of North Carolina at
schools to select evidence-based programs,
Chapel Hill.
but it does not specifically address or support
implementation or evaluation, postselection.
Learning to Developed by researchers at the Carnegie Foundation Bryk, A. S., Gomez, L. M.,
Improve for the Advancement of Teaching, this methodology Grunow, A., & LeMahieu, P. G.
provides a framework for the identification of areas for (2015). Learning to improve: How
improvement, improvement hypothesis development, America’s schools can get better at
and Agile-like improvement cycles. getting better. Cambridge,
MA: Harvard Education Press.
Verdict: A powerful set of tools with a strong focus
on starting local and then scaling. Also advocates
involvement of external researchers/coaches to work
alongside those implementing improvement.
Other promising approaches include the following: While there are encouraging case studies sup-
porting the effectiveness of these approaches,
• Collective Impact (Kania & Kramer, 2011), which there are (as yet) no meta-analyses that review or
focuses on establishing a centralized infrastructure, compare the impact of these different methodol-
dedicated staff, and structured processes. ogies at scale. There is a significant gap in both
• Scaling Up Education Reform (Bishop, the literature and the Visible Learning MetaX
O’Sullivan, & Berryman, 2010), which centers on database.
establishing goals, developing new institutions,
spreading reform, gathering evidence of impact, At present, the best that we can do is harvest
and creating opportunities for all stakeholders to the golden nuggets from these approaches to
take ownership of the reform. improvement and synthesize them into a practi-
• Many publications by Michael Fullan and Andy cal set of (hunch-based) principles and tools that
Hargreaves on scaling up (including Hargreaves educators can use to support more effective
& Fullan, 2012). education implementation runs.
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
9PART 2
The VISIBLE LEARNING®
G.O.L.D. Model
Overview
Each of the approaches to effective improve- What we set out to do in the remainder of the
ment outlined in the previous section has its own paper is to fashion these nuggets into an explicit
in-built “twist” or areas of emphasis. There are, improvement methodology that we call the Visible
however, some commonalities between all these Learning® G.O.L.D. Model (see Table 4).
approaches. To some degree, they each empha-
size the importance of carefully carrying out the We make no claim that G.O.L.D. is revolutionary.
following steps: It merely takes the best tools and processes that
we have (shamelessly) scavenged from the imple-
mentation methodologies outlined in the previ-
1. selecting the area for improvement;
ous section and blends these with trial-and-error
2. developing (or identifying and localizing) one insights that our teams have gleaned from direct
or more interventions to generate impact; involvement in large-scale improvement programs
3. implementing the agreed interventions, in the United Kingdom, United Arab Emirates,
using robust project management tools and Malaysia, Australia, New Zealand, Europe, and the
approaches; and Pacific Islands.
4. leveraging the scientific method to gather Our hunch is that these processes and tools will
data during implementation, to verify and be most valuable to educators that wish to embark
enhance impact. on large-scale improvement or what can be called
Table 4 The Visible Learning® G.O.L.D. Model
Phase 1 Goal Hunt Phase 2 Opportunity Sift Phase 3 Lift Off Phase 4 Double Back
Finding/diagnosing an Systematically Implementing the agreed Explicitly and scientifically
education challenge investigating mechanisms improvement strategies measuring the impact
worthy of everyone’s time for improvement to agree and agreeing where to
and effort on the best-fit approach go next
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
10Education Challenges. As Table 5 outlines, these even a whole system. These tools will also be of
are highly complex problems with severe impact use to individual schools, where they are united in
on future quality of life (e.g., employability, health, identifying and addressing an agreed whole-school
dignity, adaptability, resilience, happiness, and life education improvement initiative.
expectancy) and where stakeholders agree that they
want to make resolution their number one crusade. We now unpack the four phases of G.O.L.D. and
Approaches to resolution often involve the estab- present some practical tools that can be used to
lishment of a taskforce or guiding coalition that support successful resolution of substantial educa-
works with and across multiple schools—sometimes tion challenges.
Table 5 Defining Education Challenges
Education Challenges:
Education Improvement: highly complex and high risk,
Education Enhancement: more complex, addressed requires highly structured
addressed by educators by local groups of educators intervention via, e.g.,
as part of what they do via, e.g., professional G.O.L.D. thinking approach
in their classrooms learning community and additional resources
Perceived Some potential risks to quality Moderate risk to quality of life Significant risk to quality of
Impact of life if not tackled if not tackled life if not tackled
Confidence Data often white noise or Often long-term underlying Usually incontrovertible
in the Data statistical anomalies pattern in the data evidence
Stakeholder No agreement that a problem Agreement that the problem Strong agreement that the
Consensus exists exists but there are problem exists, but there are
differences of opinion about differences of opinion about
how to solve it how to solve it AND often
limited time
Resourcing Does not require external Benefits from external Obvious to everyone that it
and intervention intervention, i.e., help from requires external intervention
Reversal outside the system
Problem usually Problem almost never
self-correcting Problem rarely self-correcting self-correcting
Complexity Relatively easy to solve Difficult to solve Fiendishly difficult to solve—
solutions often generate
unanticipated consequences
Scale Classroom School Multiple schools
Source: Copyright © Cognition Education (2020). All rights reserved.
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
11Phase 1 Goal Hunt
Education challenges are often complex to define and fiendishly difficult to resolve. Change
initiatives are more likely to be successful when
• the benefits of change have been clearly defined,
• there is strong consensus that the education challenge needs resolving,
• there is deep understanding of the causes, and
• stakeholders explicitly map out what success looks like.
Therefore, we need to be clear before we embark on change that the goals are worthwhile to
warrant the time and investment in attempting systematic change. This means we need to
undertake a Goal Hunt.
A Goal Hunt is a comprehensive inquiry:
1.1 The inquiry starts with the clear purpose of identifying education challenges that are worth
everyone’s time and effort in solving; that is, what is it are we trying to make better?
1.2 During a Goal Hunt, education challenges are systematically detected, catalogued,
reviewed against local and global evidence, and ranked; that is, are we sure we have identi-
fied the most appropriate area to make better? And are we sure that we haven’t bitten
off more than we can chew?
1.3 Ideally, the fullest range of stakeholders participate in the Goal Hunt process and they are
given the opportunity to state their preferences and share their beliefs; that is, do we ALL
agree that solving this education challenge is the best use of our time and resources? Are
we all committed to making this better?
1.4 Stakeholders work collaboratively to build and validate causal models to diagnose and
explain WHY their selected education challenge exists; that is, do we share the same beliefs
about the cause of this education challenge? What are the different potential causal
explanations we need to keep in mind for when we move to the Opportunity Sift phase?
1.5 Stakeholders collaboratively set the (provisional) improvement goal, that is, this is where
we are now, and this is where we are going to get to, and this is the measure we are
going to use.
And if you can’t find an education challenge that most stakeholders agree is worth resolving, put
down your hunting implements and get back to the day job.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
12In the following subsections, we outline how you when looking at challenges in the educational land-
might bring a Goal Hunt to life in your context. scape is “What’s the worst that could happen if we
did nothing?” Table 6 illustrates some of the edu-
cation challenges uncovered by goal hunters in dif-
1.1–1.3 Hunting for an Education ferent contexts.
Challenge Worth Solving
The next step is to identify which (if any) education
To bring about change you must have a clear sense challenges are worthy of systematic resolution/
of what you want to change and why. At the system, remediation. One way of addressing this is to
district, or school level, the first step is to establish a employ the collective wisdom of the crowd and ask
G.O.L.D. Team that meets, explores, analyzes, and all relevant stakeholders (including educators, stu-
then sifts and sorts all the potential education chal- dents, parents, and community members) to liter-
lenges. This is akin to the way that Coach Glen Mills ally vote on which education challenge they think is
identified stride frequency and balance as the key the most important to resolve.
challenges or improvement areas for Usain Bolt.
There are two very good reasons for doing this.
The initial role of this hunting party is to find as The first is that, as Sir Francis Galton discovered
many challenges as there are worthy of resolution, when he asked the crowd at a country fair to guess
as possible. The question that we ask ourselves the weight of an ox (while individual guesses or
Table 6 Identifying Education Challenges Worth Solving
GOAL HUNT
Education Challenge Potential Consequences
What’s the worst that could
happen if we did nothing?
Teacher recruitment and retention • Reduced quality of math and science lessons
• Difficulty in recruiting qualified math and science • Fewer students opting for science/technical careers
teachers
• Higher lifetime cost of teacher training
• 50% of teachers leaving profession after 5 years
17% of graduating students not achieving the • Students are unable to progress to higher education
minimum standard of literacy
• When students enter the workforce they are limited
to routine roles that are susceptible to automation
Information and communication technology (ICT) • There will be insufficient functioning ICT equipment
equipment in the school district is reaching end of life across our schools
and needs to be replaced
• Our learners may not develop ICT skills
9% of students are not regularly attending school • Student learning is hampered
• Students do not graduate from school
• Restricted employment opportunities in adulthood
Teacher professional development is at an all-time low • Quality of teaching is significantly affected
in offerings and quality
• Student learning outcomes are significantly
hindered
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
13opinions were generally wide of the mark), the challenges against one another, it is also important
mean average of all guesses or opinions was that you look at the quality of evidence you have at
extremely close to the ox’s true weight (Galton, your disposal to make your collective rankings. Do
1907). What Galton uncovered was the collective you have any of the following conditions?
wisdom of the crowd, that is, when we combine
the mean average of everyone’s intuition, often • High Confidence, for example, a significant
it is not far from the empirical truth. The second amount of quantitative and qualitative data that
is that, in our educational context, by engaging have been collected from various sources, which
with all stakeholder groups, people will be more are mutually corroborative.
likely to buy in to the idea that the highest-ranked
education challenges are crusades worthy of sig- • Medium Confidence, for example, some robust
nificant investment of their own time and energy. data exist, but there are gaps that require leaps
We are also more likely to be able to determine of faith, or there are strong data, but they pull in
stakeholders’ belief and value processes, which different directions.
will be useful later, as we design and then imple- • Low Confidence, for example, while there is strong
ment our intervention. intuition and consensus that the problem might
exist, there is no hard supporting evidence. Are
One way of undertaking the voting exercise, if
judgments entirely based on gut reaction?
stakeholders have identified a very long list of
potential challenges, is by comparison of matched You will also want to consider whether the educa-
pairs; for example, the following box shows the tion challenges you have uncovered are related
comparison of a pair of education challenges. to outcomes or inputs. An outcomes challenge
is one where the existing education system is
% of students
9 versus ICT equipment in resulting in reduced life chances for learners
are not regularly the school district (e.g., lower employability, health, dignity, adapt-
attending school is reaching end of
life and needs to be
ability, resilience, happiness, and life expec-
replaced tancy). Whereas an inputs challenge is when
there is some change in the education operation
system that indirectly threatens the possibility
The idea is that every education challenge is com-
of an outcomes crisis down the line (e.g., chal-
pared to every other education challenge, and
lenges in recruitment and retention of teachers
the total number of votes for each comparison is
or school leaders).
then tallied and ranked. Stakeholders are asked to
choose which item from each pair that they rate has
the greatest priority for resolution. This process is
repeated until every option has been pitted against Going from Deductive
every other option. The ratings are then aggregated to Inductive
into an overall ranking of perceived importance.
The method for education challenge identi-
An important aspect of this process is that it fication that we have outlined in the text is
involves strong involvement from local stakehold- deductive. It starts with stakeholders laying
ers, thereby ensuring that whatever priority inter- down ideas and then looking at the evidence
ventions emerge clearly come from the community to see whether or not this conforms to their
that will go on to lead resolution. perception of reality. One potential down-
side to the deductive method is that it might
However, vote counting alone might lead to the
encourage stakeholders to identify superfi-
identification of inappropriate education chal-
cial or unimportant education challenges.
lenges. As you pit your list of identified education
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
14suggest that you then attempt to define the chal-
Another approach is to undertake your edu- lenge as explicitly as possible. For example, if your
cation challenge hunt inductively. This means selected challenge is that 9% of students are not
that you don’t prejudice your search for a attending school regularly, you will want to identify
worthwhile education challenge with precon- whether there are any patterns in the demographic
ceptions about what needs to be fixed and of that 9% and/or the duration of absenteeism, as
then take a public vote. Instead you go on shown in Table 7.
a data hunt. You start by looking at metric
data, undertake learning walks, and collect
1.4 Map the Causal Drivers
teacher and student voice data to identify/
diagnose a range of potential education In order to resolve an education challenge, we
challenges. You can then consider which need to develop and test hypotheses (or hunches)
of the challenges have the most severity of about the potential causes. This requires the
impact. G.O.L.D. Team to consider all the possible causal
factors that could be contributing to the education
Depending on the resources at your disposal,
challenge.
you could divide your G.O.L.D. Team into
two separate hunting parties (i.e., deductive One way of identifying the causal factors to an iden-
and inductive); compare and synthesize the tified challenge is called The Five Whys approach,
results from both at the end. which was originally developed by Sakichi Toyoda
and was used within the Toyota Motor Corporation
(Ohno, 1988). It involves asking and collecting data
If you can’t find an education challenge big enough about each hypothesized link in the causal chain.
to be worth solving, we would suggest that you dis- The box on the following page illustrates an exam-
band your G.O.L.D. Team. ple using the Five Whys approach.
But if you have identified an education challenge The idea is that you ask as many “why” ques-
for which there is near-unanimous agreement, we tions as you can think of and then use these to
Table 7 Education Challenge Breakdown Structure
Education
Challenge Breakdown Structure What Does This Mean?
9% of students • Not regularly attending is defined as • Our challenge group is 64% of 9% or 5.76%
are not regularly missing two days in any single week; of the student body.
attending and/or more than four days in any single
• Most of our nonattenders are adolescent
school term
boys from lower SES (socioeconomic status)
• Of nonattending cohort—36% only just group.
meet the threshold definition; 64% are
• We don’t know the causes/whether there
persistently not attending
are patterns in those causal mechanisms.
• 82% of nonattenders are boys between
• But given that majority of nonattenders
ages 13 and 17
share similar demographic features, it is
• 63% of nonattenders are low SES group reasonable to speculate that there may
be an overlap in the underlying causal
• Student voice collection was
mechanism.
inconclusive—students unwilling/unable
to articulate the reason for their absence
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
15The Five Whys Approach
Education Challenge: 9% of children are not coming
to school and it’s interfering with their learning Define the problem
Why is it happening?
1. Why? They don’t want to come. Why is that?
2. Why? They can’t relate the learning to their lives. Why is that?
3. Why? It’s boring and not contextualized. Why is that?
4. Why? Our teachers are not differentiating to their Why is that?
needs.
5. Why? Our teachers have not been trained and
supported to do this.
build a causal model. It is extremely likely that, verification methods you undertake might vary
depending on their different worldviews/beliefs, considerably depending on the nature of your edu-
stakeholders will come up with multiple five- cation challenge. Sources of data might include
ways. This process provides a window into global research including Visible Learning MetaX,
stakeholder beliefs, and it is important to inter- student voice, learning walks, parent inter-
rogate each model and come to a shared views, teacher observation, and locally adminis-
understanding. tered surveys.
In Figure 1, we illustrate how the collective or
At the end of the review process, the G.O.L.D.
majority view can be presented in the form of a
Team redraws their path analysis to reflect what
path analysis. However, you will notice that in the
they have uncovered from their research. This
illustration we have only peeled back one layer
finalized causal driver map will be central to the
of causation on each of these potential “whys”
crafting of interventions or solutions during the
(system-level discovery teams may find it bene-
Opportunity Sift stage.
ficial to add at least one more layer of “whys”
that directly map to the big six factors listed in
Figure 1).
1.5 Set the Quantifiable
You could also present your analysis as a fishbone Improvement Goal
diagram or an issue tree. However, what you have
mapped out is just a theory. You now need to collect The final stage of the Goal Hunt involves setting,
data to see whether your hunches about the agreeing, and locking specific and measurable suc-
causes of the education challenge are plausible. cess criteria for the selected education challenge
Table 8 shows how each potential causal variable (Table 9). As Peter Drucker once said, “What gets
can be investigated and recorded. However, the measured gets improved.”
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
16Figure 1 Example Path Analysis
No public
transportation
Student
dispositions
Parents cannot fund,
e.g., uniforms/
transport/meals
Student
Absenteeism
School environment
perceived to be
unpleasant/unsafe
Parents do not place
value on education
for their childrean
Poor-quality learning
experiences
Source: Copyright © Cognition Education (2020). All rights reserved.
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
17Table 8 Causal Driver Verification
Causal Causal Sources of Outcome of Variable
Variable Hypothesis Verification Data Verification Remains?
No public Students are not attending Student Not verified. All students live Removed
transportation school because the geolocations within 30 minutes’ walk from
transportation infrastructure schools, within the district.
Bus routes
is inadequate
Poor-quality Students are not attending Student voice Verified. Students Remains
learning school because they do not consistently reported
Lesson
experiences enjoy their classes and do that they found lessons
observations
not believe what they study unengaging. Lesson
is relevant Curriculum observations also suggested
review disengagement.
Source: Copyright © Cognition Education (2020). All rights reserved.
Table 9 Setting the Success Criteria
Current “To Be”
Situation Situation By When Measured How
9% of 97% of December • School attendance register data
students are students are 2022
• Regular attendance defined as 7 or fewer absences during
regularly not regularly
school year
attending attending
school school
Justification of Selection of “To Be” Values
Review of regional comparator data suggests that on average 3% of students regularly do not attend school. Our
goal is to achieve the same rate of attendance as regional comparators.
“Getting to G.O.L.D.: The VISIBLE LEARNING Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
18Phase 2 Opportunity Sift
Attempts to resolve large-scale education challenges are more likely to be successful when:
• interventions are explicitly designed;
• the designs draw on global and local research, rather than intuition alone;
• a range of alternative interventions are proposed, reviewed, ranked, and refined by local stakeholders;
• the review process includes analysis of “human factors” including cognitive biases and the degree to
which the proposed intervention aligns with or contradicts stakeholders’ existing worldviews;
• the selected intervention has an evidence base that suggests a high probability of success and that the
approach is adaptable to the local context; and
• stakeholders build an explicit and phased success map, with measurable improvement targets.
Therefore, before we blindly rush ahead and implement any approach or intervention, we need to review and
sift all the potential opportunities or options to select and design the approach that has the highest probability
of being effective within the local ecosystem.
This means we need to undertake an Opportunity Sift.
An Opportunity Sift is a systematic solutions-design exercise:
2.1 It starts by going back to the causal drivers that were hypothesized and validated during step 1.4 in
our Goal Hunt. The idea is to sketch out all the different opportunities and options for blocking, weaken-
ing, or reversing the causal drivers of an education challenge; that is, what are ALL the different ways
and tools we can use to fix this? And what evidence is there that these ways will work for us?
2.2 Next comes the development of a theory of improvement. This takes the best opportunity sketches
from step 2.1 and maps them out into a full delivery model that details the resources, activities, and out-
puts and how each of these contributes to solving the education challenge; that is, how are we going to
make this happen? What resources do we need? Who is going to do what?
2.3 We need to recognize that it’s unlikely we will have developed the perfect theory of improvement the
first time. Therefore, we need to explore all the ways it can be iterated and consider the human factors (aka
local beliefs and values) and whether our improvement model engages with these; that is, what are all
the different ways we could wiggle (iterate) our design and which wiggles do we think will result in
better impact? And does what we propose to do align with the way our stakeholders think and feel?
2.4 Finally, we need to develop a success map. This builds on the baseline and quantifiable improvement
goal targets you already set in step 1.5. It builds these out into a full results framework that sets short-,
medium-, and longer-term targets for each activity, output, and outcome; that is, how and when are we
going to measure whether we are on track?
At the end of the Opportunity Sift process, you will have identified the best options for improvement and built
these into a cohesive improvement initiative, ready to Lift Off.
“Getting to G.O.L.D.: The VISIBLE LEARNING® Approach to Unleashing Education Improvement” by Arran Hamilton and John
Hattie. Copyright © 2021 by Corwin Press, Inc. All rights reserved.
19You can also read
